Class IBSimilarity

Provides a framework for the family of information-based models, as described in StÉphane Clinchant and Eric Gaussier. 2010. Information-based models for ad hoc IR. In Proceeding of the 33rd international ACM SIGIR conference on Research and development in information retrieval (SIGIR '10). ACM, New York, NY, USA, 234-241.

The retrieval function is of the form RSV(q, d) = ∑ -x^q_w log Prob(X_w >= t^d_w | λ_w), where

• x^q_w is the query boost;
• X_w is a random variable that counts the occurrences of word w;
• t^d_w is the normalized term frequency;
• λ_w is a parameter.

The framework described in the paper has many similarities to the DFR framework (see DFRSimilarity). It is possible that the two Similarities will be merged at one point.

To construct an IBSimilarity, you must specify the implementations for all three components of the Information-Based model.

Component	Implementations
Distribution: Probabilistic distribution used to model term occurrence	DistributionLL: Log-logistic DistributionLL: Smoothed power-law
Lambda: λw parameter of the probability distribution	LambdaDF: Nw/N or average number of documents where w occurs LambdaTTF: Fw/N or average number of occurrences of w in the collection
Normalization: Term frequency normalization	Any supported DFR normalization (listed in DFRSimilarity)

Note

This API is experimental and might change in incompatible ways in the next release.

Inheritance

object

Similarity

SimilarityBase

IBSimilarity

Inherited Members

SimilarityBase.DiscountOverlaps

SimilarityBase.ComputeWeight(float, CollectionStatistics, params TermStatistics[])

SimilarityBase.NewStats(string, float)

SimilarityBase.FillBasicStats(BasicStats, CollectionStatistics, TermStatistics)

SimilarityBase.Explain(BasicStats, int, Explanation, float)

SimilarityBase.GetSimScorer(Similarity.SimWeight, AtomicReaderContext)

SimilarityBase.ComputeNorm(FieldInvertState)

SimilarityBase.DecodeNormValue(byte)

SimilarityBase.EncodeNormValue(float, float)

SimilarityBase.Log2(double)

Similarity.Coord(int, int)

Similarity.QueryNorm(float)

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

Namespace: Lucene.Net.Search.Similarities

Assembly: Lucene.Net.dll

Syntax

public class IBSimilarity : SimilarityBase

Constructors

IBSimilarity(Distribution, Lambda, Normalization)

Creates IBSimilarity from the three components.

Note that null values are not allowed: if you want no normalization, instead pass Normalization.NoNormalization.

Declaration

public IBSimilarity(Distribution distribution, Lambda lambda, Normalization normalization)

Parameters

Type	Name	Description
Distribution	distribution	probabilistic distribution modeling term occurrence
Lambda	lambda	distribution's λw parameter
Normalization	normalization	term frequency normalization

See Also

DFRSimilarity

Fields

m_distribution

The probabilistic distribution used to model term occurrence.

Declaration

protected readonly Distribution m_distribution

Field Value

Type	Description
Distribution

See Also

DFRSimilarity

m_lambda

The lambda (λ_w) parameter.

Declaration

protected readonly Lambda m_lambda

Field Value

Type	Description
Lambda

See Also

DFRSimilarity

m_normalization

The term frequency normalization.

Declaration

protected readonly Normalization m_normalization

Field Value

Type	Description
Normalization

See Also

DFRSimilarity

Properties

Distribution

Returns the distribution

Declaration

public virtual Distribution Distribution { get; }

Property Value

Type	Description
Distribution

See Also

DFRSimilarity

Lambda

Returns the distribution's lambda parameter

Declaration

public virtual Lambda Lambda { get; }

Property Value

Type	Description
Lambda

See Also

DFRSimilarity

Normalization

Returns the term frequency normalization

Declaration

public virtual Normalization Normalization { get; }

Property Value

Type	Description
Normalization

See Also

DFRSimilarity

Methods

Explain(Explanation, BasicStats, int, float, float)

Subclasses should implement this method to explain the score. expl already contains the score, the name of the class and the doc id, as well as the term frequency and its explanation; subclasses can add additional clauses to explain details of their scoring formulae.

The default implementation does nothing.

Declaration

protected override void Explain(Explanation expl, BasicStats stats, int doc, float freq, float docLen)

Parameters

Type	Name	Description
Explanation	expl	the explanation to extend with details.
BasicStats	stats	the corpus level statistics.
int	doc	the document id.
float	freq	the term frequency.
float	docLen	the document length.

Overrides

SimilarityBase.Explain(Explanation, BasicStats, int, float, float)

See Also

DFRSimilarity

Score(BasicStats, float, float)

Scores the document doc.

Subclasses must apply their scoring formula in this class.

Declaration

public override float Score(BasicStats stats, float freq, float docLen)

Parameters

Type	Name	Description
BasicStats	stats	the corpus level statistics.
float	freq	the term frequency.
float	docLen	the document length.

Returns

Type	Description
float	the score.

Overrides

SimilarityBase.Score(BasicStats, float, float)

See Also

DFRSimilarity

ToString()

The name of IB methods follow the pattern IB <distribution> <lambda><normalization>. The name of the distribution is the same as in the original paper; for the names of lambda parameters, refer to the doc of the Lambda classes.

Declaration

public override string ToString()

Returns

Type	Description
string

Overrides

SimilarityBase.ToString()

See Also

DFRSimilarity

See Also

DFRSimilarity